In many cases, an established reaction mechanism will not be available for the chemical system of interest, and the user will have to construct their own chemistry set. Developing a good reaction mechanism can constitute a substantial part of a research or development project. As shown in Figure 3.1: Developing a reaction mechanism is an iterative process, developing a good reaction mechanism is an iterative process, with comparisons to experimental data being a critical step of the process. After the first set of chemical data comprising a reaction mechanism is assembled, the results of a simple simulation need to be compared to well controlled experimental data. Agreement may not be good at first, indicating that reactions may need to be added, and/or reaction rates may need to be adjusted (especially those that were estimated, or that had large discrepancies in literature values) to give better fits to the experimental measurements.
This section discusses how one might approach the development of a chemical reaction mechanism. The discussion is divided into sections that match the input files that make up a chemistry set. The files used in the Ansys Chemkin software for describing gas phase and surface chemical kinetics have parallel structures. In practice, however, the information that they contain is obtained using very different approaches. This is a consequence of the fact that a great deal more experimental and theoretical data are available for gas-phase chemical reactions than for surface reactions.